Does the Rasa Framework also allow the use of the German GPT-2 model (dbmdz/german-gpt2 · Hugging Face) with the LanguageModelFeaturizer? Unfortunately, it did not work with Rasa version 3.1. I had also tried two older Rasa 2. x versions. There it did not work as well. Does anyone know more? Thanks in advance. My NLU pipeline configuration:
pipeline:
- name: WhitespaceTokenizer
- name: LanguageModelFeaturizer
model_name: "gpt2"
model_weights: "dbmdz/german-gpt2"
- name: DIETClassifier
epochs: 100
- name: EntitySynonymMapper
Among other things, the traceback says this (in Rasa 3.1):
TypeError: expected str, bytes or os.PathLike object, not NoneType
and
File "C:\ProgramData\Anaconda3\envs\tests\lib\site-packages\rasa\engine\graph.py", line 403, in _load_component
raise GraphComponentException(
rasa.engine.exceptions.GraphComponentException: Error initializing graph component for node run_LanguageModelFeaturizer1.```